3D Scene Geometry-Aware Constraint for Camera Localization with Deep Learning
This work addresses camera localization for autonomous driving and robotics, offering an incremental enhancement by integrating traditional geometry insights into a neural network.
The authors tackled camera localization by proposing a deep learning network with a 3D scene geometry-aware constraint, achieving significant improvements in prediction accuracy and convergence efficiency over state-of-the-art methods in indoor and outdoor scenes.
Camera localization is a fundamental and key component of autonomous driving vehicles and mobile robots to localize themselves globally for further environment perception, path planning and motion control. Recently end-to-end approaches based on convolutional neural network have been much studied to achieve or even exceed 3D-geometry based traditional methods. In this work, we propose a compact network for absolute camera pose regression. Inspired from those traditional methods, a 3D scene geometry-aware constraint is also introduced by exploiting all available information including motion, depth and image contents. We add this constraint as a regularization term to our proposed network by defining a pixel-level photometric loss and an image-level structural similarity loss. To benchmark our method, different challenging scenes including indoor and outdoor environment are tested with our proposed approach and state-of-the-arts. And the experimental results demonstrate significant performance improvement of our method on both prediction accuracy and convergence efficiency.